Question 4 - Avg. next neighbor degree

Consider the following example of a standard fully-connected Multilayer Perceptron (MLP), a type of artificial neural network. It has an input layer, a single hidden layer, and an output layer. In this architecture, neurons in one layer are connected to all neurons in the immediately following layer, but there are no connections within a layer or connections that skip layers. Here, the input and output layers have 5 neurons each, while the hidden layer has 10 neurons. Based on this description, would you expect the correlation exponent, μ, to be:



a) µ > 0

b) µ < 0

c) µ = 0

d) µ > 1

e) None of the above


Original idea by: Eduardo Bouhid

Comentários

Postar um comentário

Postagens mais visitadas deste blog

Question 3 – Network Flow

Question 2 - Strongly Connected Components